The present invention relates to identifying stale contents in a file system.
File systems, such as in the UNIX operating system, are commonly represented as a tree structure and may contain millions of files in thousands of directories. Directories are used to group files in whatever manner is convenient to the user.
Several methods have been used for archiving or deleting stale contents in a file system to provide space for more useful contents. The word ‘stale’ refers to contents that have not been accessed for a period of time that is greater than a pre-defined period of time. One method to address the stale contents problem includes the creation of a list of stale files, which can be complicated in a large file system. Further, deleting every stale file in a large list may leave the file system with directories in various states, from unaffected to partially emptied to those that are completely devoid of any files. It may also leave levels upon levels of directories containing directories without any files. Re-organizing such a file system can be extremely difficult.
In light of the foregoing discussion, a need exists for a manageable manner to identify stale contents in a file system. The present invention addresses such a need.